The general objective of this grant is to develop and test new statistical procedures in mathematical statistics for application in medical research. Three areas in mathematical statistics which have the potential for making additional significant contributions to biomedical science are (1) the analysis of censored data, (2) the analysis of binary data, and (3) robust statistical methods. The research will not necessarily be confined to these three specified areas as other areas of statistical theory may become relevant during the course of the grant. Development means theoretical work to study the properties and value of the techniques for medical applications and to adapt where necessary the techniques for practical use with medical data. Testing the procedures on actual data sets will exhibit their degree of usefulness for medical research. This grant proposal contains descriptions of a series of large medical projects with which the Biostatistics Division is very much involved and whose data sets can be used for, and will benefit from, testing new procedures. The results of this research will be communicated to the biostatistical community through the publication of technical reports. These reports will be mainly either presentations of specific research results or survey papers summarizing and discussing general progress in a certain area.